Ladder circuits are a member of a class of spatially interconnected systems and can be considered as two-dimensional (2D) systems where information is propagated in two separate directions, i.e., in time and space rep...
详细信息
Ladder circuits are a member of a class of spatially interconnected systems and can be considered as two-dimensional (2D) systems where information is propagated in two separate directions, i.e., in time and space represented by a node number. In previous work, results on the modeling, stability analysis and stabilization of such systems were given but in some cases a descriptor-like system model can result and hence these earlier results cannot be applied. This paper gives new results on the stability and stabilization of examples described by such a descriptor system model.
Discrete linear repetitive processes operate over a subset of the upper-right quadrant of the 2D plane. They arise in the modeling of physical processes and also the existing systems theory for them can be used to eff...
详细信息
Discrete linear repetitive processes operate over a subset of the upper-right quadrant of the 2D plane. They arise in the modeling of physical processes and also the existing systems theory for them can be used to effect in solving control problems for other classes of systems, including iterative learning control design. This paper uses a version of the Kaiman-YakubovichPopov (KYP) Lemma to develop new linear matrix inequality (LMI) based stability conditions and a control law design algorithm in the presence of polytopic uncertainty in the process model. The new algorithm results in a static output feedback control law that ensures robust stability along the pass and allows control requirements to be enforced over finite frequency ranges. Furthermore, a frequency-partitioning approach can lead to less conservative conditions for robust control. A numerical example to illustrate the application of the new design algorithm concludes the paper.
This paper considers the design of iterative learning control laws for systems whose state-space model matrices are functions of a vector of varying parameters. The repetitive process setting is exploited to develop a...
详细信息
This paper considers the design of iterative learning control laws for systems whose state-space model matrices are functions of a vector of varying parameters. The repetitive process setting is exploited to develop a linear matrix inequality based procedure for computing gain-scheduling feedback and feedforward (learning) controllers. These controllers guarantee acceptable dynamics along the trials and ensure monotonic convergence of the trial-to-trial error dynamics, respectively. A simulation example is given to illustrate the theoretical developments.
Matching 3D point clouds, a critical operation in map building and localization, is difficult with Velodyne-type sensors due to the sparse and non-uniform point clouds that they produce. Standard methods from dense 3D...
详细信息
ISBN:
(纸本)9781509037636
Matching 3D point clouds, a critical operation in map building and localization, is difficult with Velodyne-type sensors due to the sparse and non-uniform point clouds that they produce. Standard methods from dense 3D point clouds are generally not effective. In this paper, we describe a feature-based approach using Principal Components Analysis (PCA) of neighborhoods of points, which results in mathematically principled line and plane features. The key contribution in this work is to show how this type of feature extraction can be done efficiently and robustly even on non-uniformly sampled point clouds. The resulting detector runs in real-time and can be easily tuned to have a low false positive rate, simplifying data association. We evaluate the performance of our algorithm on an autonomous car at the MCity Test Facility using a Velodyne HDL-32E, and we compare our results against the state-of-the-art NARF keypoint detector.
This paper presents a novel reconfigurable framework for training Convolutional Neural Networks (CNNs). The proposed framework is based on reconfiguring a streaming datapath at runtime to cover the training cycle for ...
详细信息
ISBN:
(纸本)9781509015047
This paper presents a novel reconfigurable framework for training Convolutional Neural Networks (CNNs). The proposed framework is based on reconfiguring a streaming datapath at runtime to cover the training cycle for the various layers in a CNN. The streaming datapath can support various parameterized modules which can be customized to produce implementations with different trade-offs in performance and resource usage. The modules follow the same input and output data layout, simplifying configuration scheduling. For different layers, instances of the modules contain different computation kernels in parallel, which can be customized with different layer configurations and data precision. The associated models on performance, resource and bandwidth can be used in deriving parameters for the datapath to guide the analysis of design trade-offs to meet application requirements or platform constraints. They enable estimation of the implementation specifications given different layer configurations, to maximize performance under the constraints on bandwidth and hardware resources. Experimental results indicate that the proposed module design targeting Maxeler technology can achieve a performance of 62.06 GFLOPS for 32-bit floating-point arithmetic, outperforming existing accelerators. Further evaluation based on training LeNet-5 shows that the proposed framework achieves about 4 times faster than CPU implementation of Caffe and about 7.5 times more energy efficient than the GPU implementation of Caffe.
Access control is an extremely important and error-prone practice during web application. The emergence of NoSQL databases and the flexible data models they bring impose new challenges on the implementation of access ...
详细信息
Access control is an extremely important and error-prone practice during web application. The emergence of NoSQL databases and the flexible data models they bring impose new challenges on the implementation of access control within web applications. This paper presents Scout, a novel methodology for discovering access control vulnerabilities in existing web applications. Meanwhile (1) features of NoSQL database can be addressed and (2) neither application source code nor server-side session information from the developers is required. This paper implements a prototype of Scout, which targets MongoDB backend web applications. By automatically discovering the protocol layer in the web application stack, Scout introduces a data access operation model precisely representing the MongoDB actions performed in the web application, as well as inferring the access control policies. The prototype is shown to be able to identify comprehensive access control vulnerabilities in MongoDB backend web applications, and generate detailed report as the facilitator to manually fix the identified vulnerabilities.
Decentralized data-injection attack construction with minimum mean-square-error state estimation is studied in a game-theoretic setting. Within this framework, the interaction between the network operator and the set ...
详细信息
Decentralized data-injection attack construction with minimum mean-square-error state estimation is studied in a game-theoretic setting. Within this framework, the interaction between the network operator and the set of attackers, as well as the interactions among the attackers, are modeled by a game in normal form. A novel utility function that captures the trade-off between the maximum distortion that an attack can introduce and the probability of the attack being detected by the network operator is proposed. Under the assumption that the state variables can be modeled as a multivariate Gaussian random process, it is shown that the resulting game is a potential game. The cardinality of the corresponding set of Nash Equilibria (NEs) of the game is analyzed. It is shown that attackers can agree on a data-injection vector construction that achieves the best trade-off between distortion and detection probability by sharing only a limited number of bits offline. Interestingly, this vector construction is also shown to be an NE of the resulting game.
The speed and maneuverability of targets have been considerably increased, thus, improving the performance of the traditional guidance law is highly required for interceptor missile. Based on the concept of zero-sum g...
详细信息
ISBN:
(纸本)9781509041039
The speed and maneuverability of targets have been considerably increased, thus, improving the performance of the traditional guidance law is highly required for interceptor missile. Based on the concept of zero-sum game, a differential game guidance law is studied in this paper. The adversaries in conflict are considered as two independently controlled agents, the miss distance is considered as cost function, minimizing the cost function provides simultaneously the interceptor's optimal capture strategy and the target's optimal escape strategy. Because the target's future maneuver strategy cannot be exactly predicted, the proposed law is more in line with the actual situation. Simulation results show that the proposed law suitable for intercepting high speed maneuvering target, and, the proposed law has good robustness.
暂无评论